Was A1200 to blame for Falcon's failure?

[Predseda]

How many falcons and how many A1200s were sold?

[sokolovic]

Many many more A1200 than Falcons. I even think that Commodore sold more A4000 than Atari sold both Falcons models.

The A1200 was selling well before C bankrupcy. I even remember a line in French magazine Joystick (which weren't particularly Amiga fans) in early 1994 where they were saying that sales were unexpectedly high and surprised many publishers.

Of course everything collapsed after may 1994.

[Amigajay]

Originally Posted by Predseda

How many falcons and how many A1200s were sold?

Lots of quotes of the Atari Falcon selling 13-14k units by Sep 1993, before they sold the rights to C-lab and 4k units of unsold stock, so a maximum of 18k by the end i should imagine.

From what i can find going through old Amiga mags, UK wise of the 44k A1200’s sold in 1992, 30k were in the UK, 70k more were by end of Q2 1993, Com UK stated 46k sold in ‘summer quarter’ taking it to 146k by end of Q3 1993, and a reported 160k were sold Q4 Christmas 1993 period which were the last numbers reported (total of 306k) before the A1200 shortage in early 1994.

Other countries are harder to find numbers, Germany we had that 115k number quoted of March 1994. I’ve seen numbers low as 5-10k for some Scandinavian countries.

So my guesstimate would be 500k Commodore A1200’s sold worldwide. Then add the 50k Amiga Technologies unit sales from 95/96.

[lmimmfn]

I remember seeing early previews of the Falcon in maybe May/June 92, wow on paper the specs looked fantastic and completely trashed the new leaked AGA specs. However when u saw the price I just completely wrote it off.

I had an A500 at that state with 2.5Meg if RAM, there is now way I was going back to a 1 Meg machine and the 4Meg version was just beyond what I could afford, May as well go horrible PC at that stage.

However, when I did get my A1200 around late October 92, I loved it but it did feel it wasn't that much of an improvement over the A500 and that the Falcon really was next gen. I'm glad I did choose the A1200 but I've always wanted a Falcon but the prices are insane.

I was also really disappointed to hear recently it only has a 16bit bus which gimps its capabilities, again no mafa covered that limitation back in the day.

[oscar_ates]

What was the reason to cripple it with 16-bits bus?

[Galahad/FLT]

Originally Posted by oscar_ates

What was the reason to cripple it with 16-bits bus?

Cost cutting.

They did similar stupid things with the STe, giving it a 4096 colour pallete but then keeping the same three screen modes which really wouldn't showcase those extra colours without TIMER-B reload tricks.

Should have at least allowed the STe to have a native 32 colour mode, its like they saw a spec sheet for the Amiga but only read the first word of each spec without looking at the details.

Same problem with the improved sound on the STe, Amiga had 4 independent volumes for each channel, STe had one, which is why samples appear cut off on STe but fade out nicely on Amiga.

Cost cutting and without considering what those cuts would affect, because each of their machines just seemed to be reacting to what Commodore were doing and not having an actual design ethos of their own.

[eXeler0]

Originally Posted by Galahad/FLT

Cost cutting.

They did similar stupid things with the STe, giving it a 4096 colour pallete but then keeping the same three screen modes which really wouldn't showcase those extra colours without TIMER-B reload tricks.

Should have at least allowed the STe to have a native 32 colour mode, its like they saw a spec sheet for the Amiga but only read the first word of each spec without looking at the details.

Same problem with the improved sound on the STe, Amiga had 4 independent volumes for each channel, STe had one, which is why samples appear cut off on STe but fade out nicely on Amiga.

Cost cutting and without considering what those cuts would affect, because each of their machines just seemed to be reacting to what Commodore were doing and not having an actual design ethos of their own.

Id really love to hear some engineer that worked on the Falcon, explaining some of those design decisions. ;-)
I always thought the 16bit bus was due to its 1040 legacy. Looking at all component choices 16bit data bus feels like a pretty bad cost-saver.
The first ”Falcon” prototype was a 1040STE with a ”Sparrow” card featuring the 030, math copro and the DSP.
If it was designed from the ground up, its difficult to imagine that the data bus is what they would focus on in the cost cutting. They could have chosen 020 instead of full 030, low density floppy, no external scsi connector etc..
I guess you could use cheaper memory, but you only got 1MB and getting extra was already expensive anyway so that logic ”doesnt fly..”

[sokolovic]

Maybe they thought that putting a 16bit bus would'nt have been noticed. People generally checks only the basics specs, especially at that time (see post #64).

[eXeler0]

Originally Posted by sokolovic

Maybe they thought that putting a 16bit bus would'nt have been noticed. People generally checks the basics specs, especially at that time (see post #64).

Or maybe the Jaguar engineers ”did the math” for the Falcon team ;-)
32(cpu)+16(data)+24(address)=72 , see the Falcon is 72-bit, thats plenty fellas

[Thorham]

Originally Posted by TCD

There weren't many. If you fancy a trip back to the beginning of the internet here's a list: http://christophe.bray.free.fr/infor...lcon_jeux1.htm

Good lord what a waste of a nice machine

[TCD]

Originally Posted by Thorham

Good lord what a waste of a nice machine

Yep. Even on Atari forums it's hard to find info about newer games (most seem to be for Atari STEs that also run on the Falcon). PiStorm might come to the Falcon, so maybe then we'll see what a PiStormed Falcon vs a PiStormed A1200 can do

[rexsu]

Originally Posted by oscar_ates

What was the reason to cripple it with 16-bits bus?

Because the cost of RAM had increased significantly in the early 90s.
Atari had no choice!
If you want fast access to RAM with Motorola's 32-bit CPUs like the 68020/30, you're limited on memory configurations, you can't do what you want.
With a 32-Bit BUS, it's 2 or 8 MB, and with a 16-Bit BUS, it's 4 or 16 MB.
8MB would have been too expensive, and 2 MB too little for a Falcon.
We tend to forget that all the Falcon's video and audio data take up twice as much memory space as on the Amiga (16Bit sound and GFX 16 Hi-Color).
So Atari chose a stock memory configuration with 4MB and a 16-bit BUS.
I think Atari has found the right compromise, especially as the memory bandwidth with its 16Bits BUS is higher than that of the 1200 with its 32Bits BUS!!!!
Not to mention that those who wanted more performance could buy an Amiga-style accelerator card with CPU+FAST-RAM without going through the motherboard and therefore have a 32Bit BUS.

I think Atari has done a good job on this machine, especially since the video component accesses RAM in 32Bits and is much faster than AGA!
With a 16MHz BUS and a 32MHz video chip, the Falcon stock can offer much higher resolutions than the 1200, (>1600x600 on an RGB) or even SVGA 800x600 in 256 colors!!!

[rexsu]

Originally Posted by DanyPPC

Is it a joke ?

Ok, Atari Falcon is a great machine, but A1200 can be expanded, has a real palette of 24 bit, 262000 colors on screen in HAM8, PAULA DAC offers better sound than Yamaha... etc...

The Yamaha in the Falcon isn't the chip that handles the sound output, it's connected to a matrix like other elements of the machine (DSP, RAM, sound input, codec, etc.).
8 channels PCM sound uses a 16-bit codec, not 8-bit like Paula, and the signal-to-noise ratio is better too. You can find comparisons on youtube that show a clear difference in favor of the Falcon.

[lmimmfn]

Originally Posted by rexsu

Because the cost of RAM had increased significantly in the early 90s.
Atari had no choice!
If you want fast access to RAM with Motorola's 32-bit CPUs like the 68020/30, you're limited on memory configurations, you can't do what you want.
With a 32-Bit BUS, it's 2 or 8 MB, and with a 16-Bit BUS, it's 4 or 16 MB.
8MB would have been too expensive, and 2 MB too little for a Falcon.
We tend to forget that all the Falcon's video and audio data take up twice as much memory space as on the Amiga (16Bit sound and GFX 16 Hi-Color).
So Atari chose a stock memory configuration with 4MB and a 16-bit BUS.
I think Atari has found the right compromise, especially as the memory bandwidth with its 16Bits BUS is higher than that of the 1200 with its 32Bits BUS!!!!
Not to mention that those who wanted more performance could buy an Amiga-style accelerator card with CPU+FAST-RAM without going through the motherboard and therefore have a 32Bit BUS.

I think Atari has done a good job on this machine, especially since the video component accesses RAM in 32Bits and is much faster than AGA!
With a 16MHz BUS and a 32MHz video chip, the Falcon stock can offer much higher resolutions than the 1200, (>1600x600 on an RGB) or even SVGA 800x600 in 256 colors!!!

That to me doesn't seem to be a valid point, yes bus size affects RAM, but they chose a 1 and 4 Meg configuration vs a 2 and 8 Meg configuration. The 1 Meg config was a waste of time a 2 Meg config would have been a lot better and 8 Meg would have been good for professional use, the system would be more balanced with 2, 8 Meg options.

[rexsu]

They didn't choose 1 or 4 MB, but 4 and 16MB.
There were no 1 MB config in the stores when they came out, the Falcons were either 4 or 16 MB, despite some advertising to the contrary, Atari didn't want to sell Falcon 1MB.
With 2 MB, MultiTOS can't be used, there's virtually no RAM left to do anything else useful.
With 1 or 2 MB, in audio alone, you can't do anything if you're processing samples at 49KHz in 16 Bits.
Atari also wanted to make it a professional DTP machine in Germany (coz Calamus), and 8 MB wasn't enough. In fact, some people preferred the TT because it allowed configurations beyond 16MB !

A year later, 1 MB config was released by certain retailers to enable customers who wanted 16MB to buy daughterboards from third-party manufacturers (with support for SIMM modules), which enabled them to significantly reduce costs compared with Atari's memory cards!

[dlfrsilver]

Originally Posted by Predseda

How many falcons and how many A1200s were sold?

18.000 Falcons for 144.000 A1200.

[dlfrsilver]

Originally Posted by Galahad/FLT

Nope, Atari were mainly the reason why the Falcon failed.

By the time the Falcon was released, the majority of software publishers and developers had already stopped writing software for the Atari ST/e computers, and to compound the misery, those same developers would have known about the Falcon before they stopped writing for the ST, so even its imminent arrival wasn't about to contribute in any meaningful way.

One reason i heard about the publishers was that they never forgave Atari what they did with the Power Packs they gave with the ST/F/M/E computers.
Those were retail games with retail games, and Atari gave them with their computers at ridiculous prices. So when the Falcon was launched, they said in substance "We will see if Atari sell thousands of Atari Falcon, but until now, we won't do anything on it". I think they heard or knew about Atari Difficulties to provide Falcon computers (each Atari subsidiary in each euro countries were fighting to get them).

[quote]The price for the base Falcon was also another problem, sure the base Falcon was more powerful than a base A1200, the base Falcon was significantly more expensive than an A1200, and getting awfully close to a PC for price.|/quote]

This and the fact that Falcons are in fact prototype machines (if you refer to Atari industrial coding for the serial numbers), not finalized computers.

The A1200 was also released at the height of the Amiga 500's success, and pretty much EVERY publisher were onboard with supporting the A1200 from the start, the Falcon had Cubase make some press releases and Imagitec said they were working on some demos.

Basically what i said. Publishers had the sales numbers : more amiga in the wild, and more sales of softwares.

The A1200 certainly had some effect on the Falcon, I was one of those considering buying a Falcon when it was announced, but was put off immediately by the price point and that there wasn't going to be any killer software for it that made it an essential purchase.

The public didn't know what to do with such a computer. Only original Atari owners bought this computer in the end.

But the main reason was Atari, at that price point, software publishers likely wrote it off instantly suggesting not many would be sold, there were far more Archimedes computers in existence and they didn't write software for that either, so that gives you an idea of just how successful devs and publishers thought it was going to be.

Very good point. Publishers were not keen to do Archimedes dev (too few machines, even less than Atari ST anyway).

I think it was only places like France where the ST was a better seller than the Amiga that got any traction as the French devs and publishers thought that the Falcon would be the natural upgrade progression for those ST owners, hence the likes of Silmarils and other French publishers supporting Falcon initially.

A little remark regarding this fact : There were more Atari computers sold in France, with less Amiga, but the very funny fact is that more softwares were sold on Amiga than on Atari ST !

Silmarils were in the end the only publisher with Imagitec design to do commercial games on the Falcon 030.

[Pyromania]

The Jack Attack caused the Falcon to fail. And maybe the out of this world high price.

[dlfrsilver]

Originally Posted by rexsu

Because the cost of RAM had increased significantly in the early 90s.
Atari had no choice!
If you want fast access to RAM with Motorola's 32-bit CPUs like the 68020/30, you're limited on memory configurations, you can't do what you want. With a 32-Bit BUS, it's 2 or 8 MB, and with a 16-Bit BUS, it's 4 or 16 MB. 8MB would have been too expensive, and 2 MB too little for a Falcon.
We tend to forget that all the Falcon's video and audio data take up twice as much memory space as on the Amiga (16Bit sound and GFX 16 Hi-Color).
So Atari chose a stock memory configuration with 4MB and a 16-bit BUS.
I think Atari has found the right compromise, especially as the memory bandwidth with its 16Bits BUS is higher than that of the 1200 with its 32Bits BUS!!!! Not to mention that those who wanted more performance could buy an Amiga-style accelerator card with CPU+FAST-RAM without going through the motherboard and therefore have a 32Bit BUS.

I think Atari has done a good job on this machine, especially since the video component accesses RAM in 32Bits and is much faster than AGA!
With a 16MHz BUS and a 32MHz video chip, the Falcon stock can offer much higher resolutions than the 1200, (>1600x600 on an RGB) or even SVGA 800x600 in 256 colors!!!

The 16bits bus is an heritage of the 1040 STE used with a custom daughter board containing the 030, Combel and any other falcon specific chip.

That's the reason, it's not linked per se with the RAM price.

The Atari Falcon 030 is a Frankenstein computer (and prototype machines, not finalized).

Next almost all A1200 owners had a blizzard IV board with 030 and 8/16/32mb of fast RAM.

[sokolovic]

Originally Posted by dlfrsilver

18.000 Falcons for 144.000 A1200.

I highly doubt your numbers for the A1200. Considering it solds around 100 000 units in Germany alone by the end of Commodore, it would means that it just solds 50 000 for the rest of the world.
AmigaJay post seems much more accurate to me.

That said, considering how precize is your number, you maybe have some sort of source for it.

EDIT : Here you have a scan of Amiga Format from feb 1994 about sales for Q4 1993. 160 000 A1200 were sold in the UK alone in these three months.

https://pasteboard.co/Xnc5VQPKFS2D.jpg